RE: [topicmaps-comment] skills to create topic maps

[Jon Jermey & Glenda Browne <webindexing@optusnet.com.au>]

Title: RE: [topicmaps-comment] skills to create topic maps

Thomas and others,

 

Thanks for this information.

 

Six months or more ago I read 'The TAO of topic maps', and offered to write an article on topic maps for an Australian journal, Online Currents, for which I write occasional freelance articles. I then discovered that although I think I understand the general philosophy of topic maps, I don't really have a feel for the mechanics of their implementation. Hence my question about XML skills etc.

 

In the longer term I am also interested in considering the possibility of implementing topic maps, and here I also need to develop more of a feel of the practicalities. I am a freelance indexer, and one of my jobs is in the online help department of a corporation where I edit the online help index. I have also created a thesaurus and metadata for websites and the intranet within the company. There is therefore some consistency of approach to information access within the company, but there are many, many, many projects where there is no indexing-type input at all. I can see that an approach like topic maps could be of value, although I am far from having the knowledge and skills to make clear recommendations for implementation.

 

Some more questions:

 

1. People talked about companies that want topic maps already having some intellectual property in the form of a thesaurus, which could be converted to a topic map. A topic map gives additional information to a thesaurus, so presumably this all has to be added in manually (or automatically?). Eg, geographical hierarchies which are currently represented by BT/NT relationships would have to be converted to 'is part of' and 'contains' type relationships.

 

2. One of the features of topic maps is that the topic map exists outside the information resources it is applied to. However, sometimes the specific nature of the information resource is included. It seems to me that this information must be applied manually. Eg, the occurence might be a 'mention', or a 'definition', or an 'introduction'. Surely all of this has to be manually applied, meaning firstly a lot of work, and secondly, the topic map is then specific for that information resource.

 

3. The examples I have seen of topic maps concentrate on areas that are clearly defined, eg Sydney is in New South Wales which is in Australia, or X is a character in opera Y which is set in town Z. It seems to me that other very complex, not easily defined relationships won't be managed so well. Eg, there are complex relationships between the topics 'elections', 'candidates', 'nominations', 'Board of Directors'? and they don't readily fit into groups of topics as places, companies and other general categories do.

 

4. It also seems to me that topic maps which aim to be useable over many different information resources MUST lose specificity (or gain enormous complexity). I know as a book indexer that every book needs its own index, with index terms and subdivisions carefully crafted just for that specific resource. To be told to use an index from a similar book as a basis would be unimaginable. I guess using a topic map would be better compared to having to take all terms and subdivisions from a thesaurus. I guess with a quality, continually updated thesaurus this could work. And I have seen discussions on the list of whether 'more than 30%' and similar concepts could/should be topics. So maybe lots of detail is possible, but surely very complicated.

 

Thanks,

Glenda.

-----Original Message-----
From: Bandholtz, Thomas [mailto:thomas.bandholtz@koeln.sema.slb.com]
Sent: Wednesday, November 28, 2001 6:56 PM
To: webindexing@optusnet.com.au; topicmaps-comment@lists.oasis-open.org
Subject: RE: [topicmaps-comment] skills to create topic maps

Glenda,

I think the linguistic/conceptual side is the most important aspect. Even in the field of topic maps the old wisdom is valid: garbage in - garbage out.

Linguists should not be bothered with XML Schema and Syntax. We are developing an engine (KnowledgeTaxi) that will make it quite comfortable. Most of the user communities have some thesaurus or classification to start with. KnowledgeTaxi will help you to extend this existing value and to set up associations etc.

We are planning to have a first release end of Q1 in 2002. We will publish a White Paper in December.

Could you give me some more details about your project? We are very interested in real-life use-cases and community requirements.

Best regards

Thomas Bandholtz
XML Competence Center
SchlumbergerSema
Sema GmbH
Kaltenbornweg 3
D50679 Köln/Cologne
++49 (0)221 8299 264

-----Original Message-----
From:   Jon Jermey & Glenda Browne [SMTP:webindexing@optusnet.com.au]
Sent:   Wednesday, November 28, 2001 8:50 AM
To:     topicmaps-comment@lists.oasis-open.org
Subject:        [topicmaps-comment] skills to create topic maps

What skills are needed/held by people creating topic maps. I understand the
linguistic/conceptual side of it, but I am wondering about the technical
creation. Eg, do you need XML, programming skills or whatever?

Glenda

--------------------
Glenda Browne and Jonathan Jermey
http://www.optusnet.com.au/~webindexing




----------------------------------------------------------------
To subscribe or unsubscribe from this elist use the subscription
manager: <http://lists.oasis-open.org/ob/adm.pl>